Cumberland Advisors revisits the “sell in May and go away” effect. Examining the S&P’s historical performance, Cumberland notes that, after 1950, the S&P’s performance between May and October has differed drastically from its performance between November and April.
We have not yet heard a compelling explanation for why this pattern would be anything more than a temporary coincidence (other than the usual pablum about how traders are too busy schmoozing in the Hamptons), and we’ve also never seen a compelling case that the pattern has predictive value. But in any case, here’s what Cumberland found:
According to the Ned Davis (NDR) database, starting in 1950, $10,000 invested in the S&P 500 Index every May 1st and then liquidated every October 31st would only be worth $10,026 today. That’s right: had you stayed out of the stock market from November through April and only been in the market from May through October, you would have had no change during the last 57 years…
Consider the results of the reverse strategy. Buy the S&P 500 Index on November 1st and sell all your stocks on May 1st. The outcome is dramatically different. Your original $10,000 would now be worth $372,890 as of April 30th closing prices in 2008. Out of the 58 periods you would have had positive results in 45 of them and negative results in only 13 years.
Correlation doesn’t imply causation and, absent a compelling explanation for the pattern, is probably temporary. This doesn’t deter Cumberland, though, and they even add another dimension to the analysis:
Our internal studies modified the Ned Davis database. We added a component in which we assessed whether the Fed was easy, tight, or neutral. We measured the Fed by what they did and not by what they said. If they were lowering rates, they were easy. If they were raising rates, they were tightening. If they were unchanged, they were neutral.
The findings were compelling. When the Fed is tightening policy between May and November the results are nearly always horrible. When the Fed is neutral the results are flat, fair, or poor. When the Fed is easy, one can ignore the negative summer seasonal influences and remain invested.
So mightn’t the “Fed tightness or looseness” be the more relevant variable? Whether it’s May or June, raining or sunny, cold or hot, seems incidental.